Andrew Yang says, "Yes, Robots Are Stealing Your Job" in an op-ed at the New York Times. Paul Krugman thinks they're not and advises, "Democrats, Avoid the Robot Rabbit Hole." This is, of course, a classic case of asking the wrong question.
The real question is: will robots burn down your house and kill your grandchildren? Let's imagine that all those self-driving trucks and the computers needed to guide them will run on electricity generated by wind turbines and solar panels. Will the robots in the truck factories and the robots in the computer factories also run on wind and sunshine? How about the robots in the wind turbine factories and the solar panel factories and so one ad infinitum? I know an old lady who swallowed a fly...
Let's assume that it is feasible to phase out all current fossil fuel consumption by 2050 and replace it with renewable, zero-carbon energy. Does that mean it is equally feasible to provide the additional energy needed to run all those job-stealing robots? Or to put the question in proper context, would it be feasible to do it without an uncorruptable, omniscient global central planning authority?
The hitch in all this robot speculation is a little paradox known as Jevons paradox conjoined at the hip, so to speak, with it's counterpart, "Say's Law." The former paradox says that greater fuel efficiency leads to more fuel consumption, the latter paradox tells us that labor-saving machines create more jobs than they destroy. Here are two inseparable positive feedback loops that together generate an incongruous outcome. "Yes the planet got destroyed. But for a beautiful moment in time we created a lot of value for shareholders." Or lots of jobs, jobs, jobs. Or a monthly $1,000 payment to every adult "so that we can build a trickle-up economy," Choose your poison.
There is, they say, "a certain quantity of work to be done." Who says that? Good question. In the beginning, it was the political economists -- even proto political economists -- who said it. But around 1870 economists realized that the maxim conflicted with other things they had in mind so instead of professing it they began to condemn it and to attribute the idea to others -- to Luddites, Malthusians or Lump-of-Laborers. The idea that a people could always do more work was just too great a temptation. In principle, the amount of work that could be done is infinite! The robots will not replace us! The robots will not replace us!
What this job-stealing robot debate is really all about is an economics version of theodicy. "Why does evil exist if God, the creator, is omnipotent, omniscient and good?" This theological question is echoed in the puzzle about poverty in the midst of plenty and in Mandeville's "Fable of the Bees," where private vices promote public virtues. If it seems like robots are stealing your job, have faith, all is for some ultimate purpose in this best of all possible worlds, as Candide's tutor Dr. Pangloss would assure him.
Taking the Panglossian philosophy into account, it becomes clear that both Andrew Yang and Paul Krugman are on the same page. They are just reading different paragraphs. Although they disagree on what the solution is, they agree that there is a solution and it doesn't really require a fundamental change in the way we think about limits to the "certain quantity of work to be done."
4 comments:
I read this post carefully twice, and though I understand the criticism of Yang, who Dean Baker alos criticizes, I have no idea why Krugman is being criticized. Please explain.
Krugman makes the standard argument that with the right policies there is no necessity that robots (or trade) will eliminate jobs. This is true. But it amounts to "if we had some bacon we could have some bacon and eggs -- if we had some eggs." The truth of the conclusion is contained entirely in the premise. This is known as "begging the question" or "circular argument."
Just as Yang does, Krugman argues robots and employment outside of the context of climate change and the need to rapidly reduce and then eliminate net carbon dioxide emissions. So we have problem one and problem two. We solve problem one with a policy that makes problem two worse and we solve problem two with a policy that makes problem one worse. That is why I mention that the two problems are conjoined. You can't solve each one of them separately.
That was just the explanation I needed. Now the post reads clearly to me. Excellent and important, and I agree completely. Now to go back and think freshly about the problems you pose.
Also, do send this along to Dean Baker.
Post a Comment